基于集成学习的多自由度机械臂结构参数优化研究  

Research on Optimization of Structural Parameter of Multi-dof Mechanical Robot Based on Ensemble Learning

作　　者：钟明灯[1] 童慧芬[1] 李岩 ZHONG Mingdeng(Minnan University of Science and Technology,Quanzhou Fujian 362000,China)

机构地区：[1]闽南理工学院,福建泉州362000

出　　处：《长春工程学院学报（自然科学版）》2023年第1期81-85,共5页Journal of Changchun Institute of Technology：Natural Sciences Edition

基　　金：泉州市科技计划项目(2020C068);福建省中青年教师教育科研项目(JAT210499)。

摘　　要：在多自由度机械臂运动中存在关节转角范围小、角速度波动大的问题,严重影响到了机械臂的工作效率与质量。为了解决该问题,提出了基于集成学习的多自由度机械臂结构参数优化方法。通过定义多自由度机械臂末端执行器的位置坐标,计算出机械臂末端执行器的线速度,根据机械臂末端执行器的角速度与关节角之间变化量的关系,构建了多自由度机械臂运动学方程,通过计算机械臂运动方程的正解和反解,分析了多自由度机械臂动力,结合集成学习算法对机械臂运动轨迹的记忆学习,并根据机械臂各关节的转角范围和约束条件,构建机械臂结构参数优化函数,实现了机械臂结构参数的优化。试验结果表明,这种方法不仅可以通过扩大关节转角的范围来保证机器人运动的稳定性,还可以通过降低关节角速度的波动来提升结构参数优化效果。实际应用效果好。In the movement of the multi-dof mechanical robot,there are problems with small range in joint angle and large fluctuation in angular velocity,that seriously affect the working efficiency and quality of the mechanical robot.In order to solve these problems,a method of optimization on structural parameter of the multi-dof mechanical robot based on Ensemble Learning is proposed.It calculates the linear velocity of the end effector of the mechanical robot by defining the position coordinates of the end effector of the multi-dof mechanical robot,and establishes kinematic equations of the multi-dof mechanical robot according to the relationship between the angular velocity and the joint angle of the end effector of the multi-dof mechanical robot.It analyzes the dynamic of the multi-dof mechanical robot by calculating the normal and inverse solutions of kinematic equations,constructs the optimized function for the structural parameter of the multi-dof mechanical robot while combining Ensemble Learning algorithms with memory learning of the motion trajectory of the mechanical robot and based on the range of joint angle and constraint conditions of each joint of the mechanical robot.It is indicated that the method can not only ensure the stability of robot’s motion by expanding the range of joint angle,but also improve the optimized effect of structural parameters by reducing the fluctuation of angular velocity of the joint,which is of good practical use.

关 键 词：动力分析 集成学习 参数优化 机械臂结构 多自由度 运动学分析 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]